Data Centre Decommissioning Collection

Decommissioning a rack is not the reverse of installing one. Installation happens into an empty, prepared space on a planned date; decommissioning happens around live equipment, under change control, often in a facility where the vehicle cannot get near the floor and where every item leaving has to be accounted for. Add a concentration of storage media far higher than any office produces, plus batteries that classify differently from everything around them, and the removal becomes the constrained part of the project rather than an afterthought.

What makes data centre decommissioning different from an office clearance?

Density. A single rack concentrates more weight, more storage media and more access constraint than a floor of office equipment, and all three change how the removal has to be planned.

An office clearance is a volume problem — lots of light items occupying space. A decommissioning is a weight and access problem: populated racks and UPS units are heavy, awkward to manoeuvre, and frequently sited where a vehicle cannot reach without a long internal route. On top of that sits a data concentration that an office never approaches, since every drive in every server is a data-bearing asset requiring individual accounting. Whereas an office refresh can tolerate an approximate asset list, a decommissioning cannot.

What equipment comes out of a decommissioning?

Four broad categories, and they do not all classify the same way. The split matters because hazardous and non-hazardous items cannot travel on the same transfer document.

UPS batteries are the item most often overlooked in planning, because they are treated as part of the infrastructure rather than as equipment being disposed of, right up until the point someone asks how they are leaving the building.

How are data-bearing assets accounted for?

Individually, against a list built before anything moves. A single populated 42U rack can hold several hundred drives, which makes a quantity-based handover useless as evidence.

The list should record serial numbers at drive level where drives are being removed separately, or at chassis level where whole servers travel intact with their media installed — but the two approaches cannot be mixed without saying which is which. At collection the items are checked against the list and the handover signed by both parties, fixing the point custody transferred. Everything downstream depends on it: a certificate of data destruction listing four hundred serial numbers is only meaningful if there is something to reconcile it against. For custody-critical projects see chain-of-custody IT collection.

What access constraints affect the removal?

Route, not just door width. The limiting factor is usually the internal path from the floor to the vehicle — lift capacity, ramp gradients, door swings, floor loading and the distance equipment must be moved by hand.

Goods lift capacity is the common hard constraint, since a populated rack can exceed what a passenger lift will take, forcing dismantling on the floor rather than at the loading bay. Time-of-day restrictions add a second layer: many facilities permit removals only outside core hours or within booked loading windows, which converts a straightforward job into a sequence of short attendances. These constraints should be walked and confirmed before the collection is scheduled, because discovering a lift limit on the day stops the work entirely.

How does decommissioning interact with change control?

It sits inside it. Removal of equipment from a live facility is a change like any other, and the collection has to fit the approved window rather than the other way round.

This is the practical reason decommissioning collections are rarely same-day. The window is agreed in advance, often narrow, and frequently outside normal hours — so the value of a carrier is predictable attendance within that window rather than speed. It also means the classification and paperwork should be complete beforehand: a load that needs sorting into hazardous and non-hazardous during a two-hour overnight window will not finish. Where equipment is going into service rather than out of it, see data centre server deployment.

What documentation does a decommissioning produce?

The same set as any disposal, but at greater scale and with more items requiring separate classification. The transfer document, an itemised asset list, destruction evidence and a named treatment destination.

The transfer note carries the waste description, EWC code, quantity, containment, place and date, both parties' details, the carrier's registration number and your SIC code, signed by both sides and retained for two years. Where batteries or other hazardous items are in the load, those move on a consignment note instead. Only an approved authorised treatment facility can issue WEEE evidence notes into the national system, so the treatment destination is worth recording against the project. Full detail in our IT disposal documentation guide.

Can decommissioned equipment be resold or reused?

Frequently, since equipment retired on a refresh cycle rather than through failure often has years of service left. Reuse is the better environmental outcome and can offset project cost.

The data position is what governs it. Equipment leaving in working order carries readable storage unless it has been erased to a verified standard first, and in a data centre context that means several hundred drives, each of which has to be individually accounted for. Unlike physical destruction, where the media ceases to exist, a reuse route ends with functioning drives in somebody else's estate — so the erasure evidence is the whole of the control. Decide the route per asset class before the project starts, not per item during the removal.

Should the asset list be built at drive level or chassis level?

Either works, provided one is chosen and applied consistently. The two produce different units of reconciliation, and mixing them without saying so is what makes a large list unusable.

Chassis-level listing is lighter work up front and adequate for most decommissioning, but it pushes the itemisation onto the processor — so confirm before collection that they will provide it, rather than discovering afterwards that the certificate counts chassis and your risk register counts drives.

What should be confirmed before the removal window?

Six things, all physical or documentary, none of which can be resolved during a two-hour overnight slot. A decommissioning that stops halfway is almost always stopped by one of these.

1. Goods lift capacity and whether a populated rack exceeds it, since that determines whether dismantling happens on the floor.
2. The full internal route — door swings, ramp gradients, floor loading, and the hand-carry distance to the vehicle.
3. The approved change window and any site access restrictions on hours or loading bay booking.
4. The hazardous split, particularly UPS batteries, which may need consignment-note handling.
5. The asset list, at drive level or chassis level, stated consistently.
6. The treatment destination, named and recorded against the project.

Walking the route in advance sounds excessive until a lift limit discovered at 22:00 ends the window with half a suite still racked.

How is a partial decommissioning handled?

As a removal from a live environment, which is materially harder than clearing an empty suite. Equipment being taken out sits alongside equipment that must keep running, so the constraint is protecting what stays rather than moving what goes.

Whereas a full decommissioning can treat the floor as a work area, a partial one operates in aisles with live power and cooling, under change control, with a defined boundary between what is in scope and what is not. That boundary needs to be unambiguous on the asset list before anyone arrives — identified by rack position as well as serial number — because an item removed in error from a live estate is a production incident, not a paperwork problem. Where equipment is being installed rather than removed, see data centre server deployment.

How does T&C Logistics support a decommissioning?

By attending within the agreed window on dedicated vehicles, with the documentation prepared in advance. We carry under Environment Agency registration CBDU654368, with treatment at a partner-operated approved facility.

Transport runs under our Operator's Licence with GPS tracking and goods-in-transit cover; loads are dedicated rather than consolidated, which is what keeps a large itemised consignment identifiable end to end. Batteries and other items classified as dangerous for carriage are handled under our hazardous goods procedures by ADR-licensed drivers. We do not operate a treatment site and do not issue destruction certificates — those come from the parties that perform that work. For the wider service see IT equipment collection and disposal, or get in touch to plan a removal.